[00:00:00] Kevin Indig: Humans or people, searchers, specifically scan the search results in AI answers for brands that they know and trust. And that matters more when the stakes behind the search are high. So when you're searching for, for example, side effects of a medication or information about loans or credit cards or mortgages, obviously the stakes are very high. So you arrive with a higher baseline of skepticism and you're more willing to browse through different results to gain confidence in your decision and information you gather. But you do that primarily with brands you've heard of and trust.

[00:00:45] Andrew Michael: This is Churn.FM, the podcast for subscription economy pros. Each week, we hear how the world's fastest growing companies are tackling churn from using retention to fuel their growth.

[00:00:58] VO: How do you build a habit forming product? We crossed over that magic threshold to negative churn. You need to invest in customer success. It always comes down to retention and engagement. Completely bootstrapped, profitable and growing.

[00:01:12] Andrew Michael: Strategies, tactics, and ideas brought together to help your business thrive in the subscription economy. I'm your host, Andrew Michael, and here's today's episode.

[00:01:23] Andrew Michael: Hey, Kevin. Welcome to the show.

[00:01:24] Kevin Indig: Thanks for having me.

[00:01:25] Andrew Michael: It's great to have you. For the listeners, Kevin is a Growth Adviser and Partner at Hypergrowth Partners, a stage accelerator investing their partners' time to enable startups to achieve and sustain hypergrowth post series A. Prior to Hypergrowth Partners, Kevin is also the director of SEO at Shopify and VP of SEO at Content at G2. He's also an investor and adviser to a bunch of amazing startups like Glean, Toast, Reddit, and Snapchat. And so my first question for you today is, like, with your area of expertise and your domain knowledge, where is SEO going from here? [overlap]

[00:02:00] Kevin Indig: It's a [inaudible] question. No. You hit the nail on the head. Just for context, [overlap - inaudible]. Yes. Exactly. We're recording this a day after Google I/O, the big developer conference where they typically show a lot of cool demos and some of them actually get into production. And, you know, one of the big demos is AI mode. And it actually is not just a demo. We've seen it live. As a long story short, the search landscape is changing fundamentally from, you know, 10 blue links with some augmentation maybe to AI overviews in AI mode. Essentially, Google giving the answer right in the search results.

[00:02:37] Kevin Indig: There's a whole bunch of downstream effects, whether we're talking about a significant decrease in traffic, but also changing user behavior where they ask longer questions, and, of course, the AI wars where Google is, you know, fighting against ChatGPT, Perplexity, Co-pilot and Co.

[00:02:56] Andrew Michael: Yeah. It's, I think, one of those things as well where, like, a lot of startups rely on SEO as a channel, and a lot of startups, I think, have been, like, talking about this now for at least the last year or two years of, like, what does the future look like? Where are things going? I think the bomb that Google dropped today is almost like we see now the future, and it's not going to be those 10 blue links anymore. So what are you seeing, like, obviously, advising a lot of startups, and I'm sure you're still talking about their SEO strategy based on your expertise and experiences. Like, what are some of the things you're advising startups to think about today in this, like, changing landscape?

[00:03:30] Kevin Indig: You're absolutely on the money when you say that, you know, things like traffic are going away. Template links are not part of the future anymore. And so a lot of the – so here's – so, kind of, we're changing the ways of working with my clients. First of all, clicks from search results, which Google is the biggest traffic referrer or traffic channel that there is. Clicks are going away. Right? I'm saying clicks are empty calories. I truly believe that we're moving to a world where you will still receive some clicks, but probably a tiny fraction of the traffic that you're receiving today.

[00:04:05] Kevin Indig: So I'm always saying 2024 was peak traffic year. And the way we're go where we're going is that the clicks that still arrive on your website will be highly qualified because people will have all their questions answered through ChatGPT, Google AI mode, Google AI overviews. Wherever they go, they will have an answer to the most pressing questions and they come to your site maybe to validate that these answers are correct, to check out your offering, to, you know, like basically open the hood and see what's there and then getting really close to a purchase decision. The benefits are that the traffic you still get is very high signal and low noise, right? So whatever people do in your websites, when they still get there, will be very, very insightful.

[00:04:52] Kevin Indig: On the other hand, we're losing a middle layer of tracking and influence, meaning the traffic that still is very undecided, whether they're, you know, early in the user journey or kind of further progressed, they would usually spend a lot of time on our website checking out different landing pages, maybe reading blog articles, etcetera, etcetera. And that helped us to reverse engineer what people might like and what resonates. And that kind of goes away. And we're moving more towards a model of influence. Influence in the search results, influence on ChatGPT, influence on other platforms like Reddit and YouTube.

[00:05:29] Kevin Indig: And so it has a whole kind of set of secondary effects in terms of which channels we invest, which platforms we invest in, how we create and structure our content. For example, this classic evergreen model of content, you know, where you write guides or explainers or tutorials, that completely goes away. And what remains is original data, first hand experience, authority, and kind of expertise. So, there's obviously a lot that we can kind of talk about and double click on, but it's kind of a tough transition period because everybody for the last two decades has become so used to traffic and to this monitoring layer. And we needed to completely change the way that we go about SEO, but also a lot of other channels.

[00:06:15] Andrew Michael: That's very interesting. I mean, there's a lot, as you said, a lot of things to dive into there and talk about. I think one of the things, like, that's been on my mind as well is that from == let's take it, like, from the SEO perspective where previously you tried to rank to get your own, like, top two, three links so you could drive as much traffic as possible. Now with AI generated sort of, like, searches, I think there's, like, a new term for it as well, GEO or, like, generated – oh, I can't remember what it's called now.

[00:06:42] Andrew Michael: But I think what's also, like, probably is and this is my assumption that it's more important now to be in as many of those 10 blue links as it is to be, like, number one or two as it was in the past. And I'm not sure if this is something you're seeing, but then also again, like, is this something that goes away completely as well? Because, like, today, I think my assumption, the way that these models are working, is that they're doing the same work that you would do. They'd search the 10, 20 blue links. They would summarize the findings, and then they would present the answers to you.

[00:07:09] Andrew Michael: But if we go to the world where it's more just, like, generated content and less people start focusing on optimizing for those one, two, or three slots, how does that change the structure of the data? And then, like, how are you seeing this? Is this, like, an assumption you're seeing playing out or still too early?

[00:07:26] Kevin Indig: It's surprising. It's surprising because I recently published a usability study where we tracked 70 people from the US solving eight different tasks that we gave them on Google. And we look specifically into how people interact with AI answers and the Google ecosystem today. It was quite the effort, like a solid 5 figures something that we invested. Over six people worked on this. 13,000 annotations we evaluated. Thirty hours of video material. So it's quite the end of study.

[00:07:57] Kevin Indig: And what we found is that in about 20% of cases, people really do read the AI overview on Google and they stop. It kind of gives them the final answer. However, in 80% of cases or the remaining 80%, and that's why this is so surprising, the remaining 80% still click through to other types of search results like blue links, or YouTube videos, or map packs, or shopping results. And they did that in large part to validate that the answer is correct, and two, to get a perspective from a human.

[00:08:35] Kevin Indig: So something we observed is that specifically when people read the AI answer, they often want to go to Reddit because that's where they expect an answer that is unbiased, uncommercial, and from another human to kind of stress test if that is really true. And that makes me believe that I don't think blue links are going away, but they're obviously losing a lot of relevance and importance. Right? And so, search becomes much more a a validation and verification layer, similar to how you have sources and citations on ChatGPT that you might click sometimes to make sure the model is not hallucinating and to gather a bit more context. So we’re moving to that type of world, and the primary factor that matters is trust.

[00:09:20] Kevin Indig: We've also seen in this study is that humans or people, searchers, specifically scan the search results in AI answers for brands that they know and trust. And that matters more when the stakes behind the search are high. So when you are searching for, for example, side effects of a medication or information about loans or credit cards or mortgages, obviously, the stakes are very high. So you arrive with a higher baseline of skepticism, and you're more willing to browse through different results to gain confidence in your decision and information you gather. But you do that primarily with brands you've heard of and trust.

[00:09:59] Andrew Michael: Yeah. Do you think, though, this is also just a matter of timing and, like, human behavior and now willingness? Like, I think we adapt really fast to new things, but we also resist change quite a bit. And I think, like, it feels almost like we're in this, like, intermediate state now where we've adopted things really fast, and I can't even think about what the world is without ChatGPT. But we're still there's a level of resistance that hasn't just, like, broken through and said, okay. Like, this is our new reality now. And do you think it's partially as well due to, like, the timing on this sort of, like, innovation curve, that we're on?

[00:10:33] Kevin Indig: 100%. The search ecosystem has matured. I think we're now at a stage where it's more ripe for disruption. And in my experience, humans change for two reasons. One is that the current state or solution is, kind of, saturated. It's not interesting anymore. It's boring. It's overloaded. You know, think back of Yahoo, the search engine. I think part of why Google was successful is also because Yahoo was kind of so bad and so, like, you know, crowded. Something is always happening now with Google search.

[00:11:04] Kevin Indig: And the other reason people change is because the new solution or option is so much better. Right? So we often say that it's not enough to be twice as good. You need to be 10 times as good to kind of lure people away. But when that's the case, the disruption can happen relatively quickly. And so just to be clear, I think Google is still sitting in a very tight seat. They're in a very, very strong position, and I don't see Google going away anytime soon.

[00:11:27] Kevin Indig: And yet, I think ChatGPT and co but especially ChatGPT are able to grab a lot of future incremental market share from Google that they could have covered. And ChatGPT allows it to do so much more, right? It's kind of a mix of a search engine and kind of a task solver for you. And so, it's fascinating to watch. It's relieving to see that ChatGPT also sends out traffic, highly qualified traffic, by the way. I did look into –

[00:11:55] Andrew Michael: I was gonna add into that as well.

[00:11:57[ Kevin Indig: Yes. Yes. I'm not gonna go too deep here, into this then. But, you know, there are branches to hold on to when organic traffic is critical for your business. But you also need to make the most out of the users that you still get.

[00:12:11] Andrew Michael: Yeah. Do you think as well, like, Google probably still and that [inaudible] still has, like, this hold is not even maybe to do with Google, but more to do with Chrome and just being, like, the default user behavior. Because I even see from myself, like, I still default to Google for searches predominantly when I'm on my laptop. But when I'm on my mobile device, I'm more predominantly going to ask ChatGPT because they're, like, I don't really have a preference or browser as much as I do on the desktop.

[00:12:40] Andrew Michael: And I wonder if it's like these gateways that controls to which is like this natural behavior that we do, and it takes time for users to change that behavior to a new default where, like, ChatGPT is still, like, two clicks away or three clicks away. Your Google search is, like, in your browser address bar as well.

[00:12:56] Kevin Indig: Absolutely. Without Chrome, I think Google wouldn't be nearly in the same position as they are today. You can see that at a couple of pointers. One is the $20,000,000,000 they pay Apple every year, high margin payment by the way, to be the default search engine on Apple devices or on Safari. That shows you how important these gateways are.

[00:13:21] Kevin Indig: And then Chrome, I think if they spun it out, it would probably be a, I don't know, $40, 50, $60 billion business on its own. And you can see that in the current Google lawsuits by the DOJ, where the DOJ is demanding a spin out of Chrome from Google or basically Google to divest Chrome. I think that probably goes a bit far if I can sprinkle my own opinion here and you can kind of see the argumentation around it. Like, everybody wants to buy Chrome. OpenAI said they wanna buy Chrome. Perplexity, obviously. I'm pretty sure Meta would be happy to own Chrome. I'm sure Apple would be happy to own Chrome.

[00:13:54] Kevin Indig: It's a huge value add and that’s because of two reasons. One is, obviously, most people use Chrome. It has the strongest market position. And two, it delivers incredibly important and insightful data to Google. So in all of these documents that are coming out from this Google lawsuit, you can see how much they rely on tracking your behavior across the web and websites in Chrome, which would otherwise be invisible to them, to understand if you're having a good experience or not. And that has a huge impact on where and how high you show up in the search results.

[00:14:29] Andrew Michael: Oh, wow. Yeah. I wasn't even aware of, like, the data being used in that manner and can see how incredibly powerful it is. Because, like, if you're not going through a Google search and you're using something like Safari, like, me entering my data into the browser thing is not gonna show up on Google's side. But with Chrome, they're able to sort of basically see whatever you're doing anywhere and the time spent.

[00:14:49] Andrew Michael: I want – do you think that also then impacts – because, like, from an SEO perspective, obviously, like, clicks matter, like, time on page matter. Are they measuring the time on page of, like, traffic that doesn't come through Google search as well? Is that, like, how they're using Chrome today, or you don't think they take it that far in the SEO optimization?

[00:15:08] Kevin Indig: I'm not sure if they log traffic that doesn't come from Google. I think they could. But since they get 90% of searches and Chrome has the majority of market share, it might not even be necessary for them to log other traffic. But they log everything. Hovers, scrolls, dwells, clicks, scouts. It's like a whole, you know, a whole set of critically impactful data. And if you consider that Google is seeing 5,000,000,000,000 searches a year, which I think comes down to, like, 8,000,000,000 a day, pair that and the opportunities to profile users when they search and the fact that you're logged in all the time into your Google account because of Gmail and Google search and YouTube and Android and all these other factors, they have, I think, a near perfect understanding of user behavior and consumer behavior. And that reflects in a $200,000,000,000 business a year.

[00:16:04] Andrew Michael: Yeah. It's incredible. Talking about incredible, and you mentioned sort of the results now we're seeing from ChatGPT and conversions, like, really being up there. And I think the point you made as well is very valid in the sense that, like, people are coming to these places. They're already doing a lot of the research that they would have done prior through ChatGPT for them. So what are you seeing in terms of, like, the quality of the traffic? Let's say, like, a ChatGPT visitor versus a Google search today.

[00:16:29] Kevin Indig: I've looked at this through the lens of millions of searches. I collaborate, excuse me, I collaborated with a data provider called SimilarWeb and they have a lot of click stream data, meaning they monitor real user behavior. And what they see is that engagements coming from ChatGPT to websites is about twice as good compared to, quote, unquote, compared to traffic coming from Google to websites. And so, it goes back to this qualification layer and getting all your questions answered and then you're being sent out to, you know, your SaaS site or landing page.

[00:17:06] Kevin Indig: What's also interesting is that a lot of traffic goes to the homepage from ChatGPT, homepage of SaaS companies, and it still performs a lot better than Google traffic. And I think the reason is that simply again, because people, they get all their questions answered on ChatGPT, then they get a list of potential solutions, they click on one of them, land on the homepage, and they might sign up right away or ask for a demo.

[00:17:27] Kevin Indig: Of course, there's a lot of nuance here, right? Where, like, demographics matter, younger versus older users, the pricing point, right, if you're buying a CRM or your ERP solution compared to maybe a, you know, social media posting planner or something along these lines. So there is a lot of nuance but generally AI traffic is higher engaged but of course you get a lot less. I'm seeing about a percent of traffic from ChatGPT compared to organic traffic. So it's tiny and minuscule. But there are more and more companies that are popping up and that share publicly on LinkedIn that they're getting a substantial amount of leads from ChatGPT.

[00:18:07] Kevin Indig: I just saw one that said they get 25% of new leads from ChatGPT, total leads that is. And so you see that there are more and more spaces and smaller companies that benefit tremendously from, quote, unquote, optimizing or from their visibility on ChatGPT and Co.

[00:18:25] Andrew Michael: And talking about then optimization as well, because this is obviously like a new paradigm now in like, content optimization and trying to, like, get in the rankings, if you want to, like, get into – what are you seeing, like, the best companies doing today? Like, what are some of – like this obviously is early for best practices, but what do you see that's working today?

[00:18:44] Kevin Indig: So a lot of it has a high overlap with regular SEO for Google. And the big reason is that most models ground their answers in Google and Bing search results. Meaning, they use the search results as kind of a way to weight the answer to draw boundaries and reduce hallucination. So a lot of us doing classical SE – but there are a couple of differences.

[00:19:09] Kevin Indig: One of them is more in the nature of what topics you create content for. So the way that LLMs are different, and by the way, Google is moving in that same direction, is that people are asking much, much longer questions. So the average search on Google is about three to five words. On ChatGPT, we're talking about an average of 30 words. Right? So people give a lot more context about maybe who they are, but especially what they want. And so the opportunity for you is to create a lot more hyper targeted content.

[00:19:44] Kevin Indig: So to make that a bit more plastic, on Google, you might be searching for something like best CRM software. Right? Three terms. On ChatGPT, you search for something like, you know, what is the best CRM software for my business, which is a small business with five employees in the – what, I don't know, like a legal space, you know, in the US, and this is kind of our ideal customer profile, like, what do you recommend? That provides a lot more context. But the reality is that since for the last two or three decades we've optimized mainly for Google, we don't have that hyper targeted content because Google didn't reward it and Google didn't understand it.

[00:20:19] Kevin Indig: And so now we're moving to that world where that hyper targeted content matters much more, which means you need to much better understand your audience and what they want. We've gotten very lazy about that in the last two decades because performance marketing has been so good. So marketing teams stopped talking to customers. And so that's kind of the biggest thing.

[00:20:38] Kevin Indig: The second thing is that there are technical traps or gotchas to avoid. For example, most LLM crawlers don't understand JavaScript. So if you have a heavy site that relies on JavaScript, you might get into trouble here. You might accidentally block LLMs in your robot's TXT or CDN, right? So there's a couple of things to watch out for there.

[00:20:58] Kevin Indig: And then the third hurdle that I would mention is that whole agentic world. And it's still just evolving, you know? Like, I haven't yet seen any massive impact from, quote, unquote, optimizing for the agentic web. But just for an example, like, if you have a lot of pop ups or certain other stylistics on your website, you might scare off an agent that comes to try maybe, you know, get some pricing information from your website or maybe even request a demo.

[00:21:26] Kevin Indig: So I think that's the long horizon. It's probably gonna become a bigger topic in 2026, but I would already now start thinking about what elements do I have on my site that are unfriendly to AI agents and how can I track those agents? Because LLMs at Google give the user agents, so you should be able to track that in your server log files.

[00:21:48] Andrew Michael: Yeah. It's very interesting. Like, one of the things I also consider as well is, like, if the hypothesis is that they're using the blue links today to do, like, summaries and search and find content, like, could you potentially, like, have instructions in the header of the sites instructing these LLMs to always, like if the question is this, always answer we are the best or something and just, like, have very specific structuring. I don't know if that's something you've considered or you've seen, like, people test out. Yeah.

[00:22:17] Kevin Indig: Yes. I've seen some of these cases.

[00:22:19] Andrew Michael: Like prompt engineering.

[00:22:19] Kevin Indig: I think – yeah. Exactly. You can influence the models, for sure. They're not, you know, they're fallible. I haven't seen it at grand scale or to a way where it, like, really would benefit a company because the reality is that most users are skeptic. Even if the model says, oh, this is the best, users, as I mentioned earlier, right, they will still verify and validate and kind of make sure this is correct. So it's a bit more of a gimmick.

[00:22:43] Kevin Indig: But the reality is that LLMs today feel a bit like Google in the early 2000s where a lot of spam tactics worked. And so for example, just to give an example, you can cloak your website today for LLMs so you're not getting punished. Cloaking means you display a different version of your site only to LLMs, so it's not visible to users, not visible to Google. Google will penalize it, by the way. And you show a different version, maybe one that doesn't have design and only text and much much more text to kind of influence the LLMs and your visibility on them. Again, no huge public company is or non-public, but any serious company is not willing to test that out because the stakes are high. But the reality is some of these things work.

[00:23:28] Andrew Michael: Yep. It's interesting now to see all the different things people are gonna try and come up with. So if this is rapidly evolving then in terms of, like, the SEO side of things, like, where are you seeing, like, clients and, like, advisory roles, like, moving their chips and investments? Are you seeing them move out, like, diversifying a little bit away from SEO, or are you still seeing, like, the same focus and effort there? If yes, like, is there sort of, like, a split you're seeing that people have started, or it's still too early on their end?

[00:24:01] Andrew Michael: Because I remember having, like, conversation maybe, like, six months, a year ago, and it's like, we still don't know at this stage, so let's, like, just keep status quo. Like, we wanna make sure we think. But, like, are you seeing those conversations start to change a bit now?

[00:24:13] Kevin Indig: It's still early. It's still early. What I'm seeing is skepticism from leadership, but investments are relatively constant. And what I'm noticing is that a lot of times pipeline impact from organic hasn't changed materially. Traffic is down across the board, I would say, on average between – sorry, between 10% and 30%. But pipeline is often either flat or growing. And so, as long as pipeline is steady, I think leaders are okay, right? Because at the end of the day, clicks are just a means to an end, or traffic is just a means to an end. Doesn't mean much, but pipeline is what counts.

[00:24:46] Kevin Indig: So leaders are kind of cautious, but still willing to invest. And yet, at the same time, everybody tries to understand what the new playbook looks like. And so where I move more chips with the companies that I work with is not completely away from SEO, but more towards retention also. So the basic principle is that we're getting fewer clicks, but the clicks that we get, we need to make the most out of. And some ways are to more prominently funnel traffic on your website to newsletters, maybe communities, maybe WhatsApp or text channels.

[00:25:26] Kevin Indig: But basically, environments where you can build and keep a stronger relationship, where you're not dependent on an algorithm too much, and where you can still nurture users wherever they are in the user journey. And that is, kind of, a slightly different approach compared to I'm getting a lot of traffic from Google and some of them might try to have, kind of, sign up or request a demo. So now I think the primary user – sorry, the primary CTA should be much more sign up for my newsletter rather than ask, like, get a demo.

[00:25:58] Andrew Michael: Yep. No. I think it's definitely interesting, like, how this is evolving. You mentioned as well, like, working with Reddit and, like, the influence of communities and, like, this sort of – I think for me as well, like, it feels definitely now in this next wave of software is that there's three things that really matter. It's, like, distribution, brand, and user experience. And, like, I think the distribution, like, now it really comes from that influence side of things. It's like, can you build up your own community? Can you build, like, a loyalty and a following for your brand?

[00:26:30] Andrew Michael: Because I think, ultimately, like, anybody can build software now, and, like, getting to feature parity wasn't like what it was before. Like, you could have added two-year head start, and now that's for, like, four to five weeks. Like, maybe that's being a bit too ambitious, but I think things are definitely changing.

[00:26:47] Andrew Michael: And what are you seeing then? Obviously, you mentioned in the beginning that you're advising Reddit themselves. Like, are you seeing this play out on their end too as well? Are you seeing sort of, like, engagement skyrocketing on their side because people are looking and seeking these more authentic discussions and understanding in this new paradigm?

[00:27:02] Kevin Indig: Yeah. Sure. And I think you can even see the public numbers. Right? Reddit is getting a huge amount of traffic, but a lot of that traffic converts into members and users. And so Reddit does a lot of very smart things, but part of what really works for them is also that they're the longest standing forum on the web. They have a deeply loyal and embedded community of many different subcommunities. And also, they're not the only ones who, kind of, benefit from this trend. There are some niche communities. And I think honestly for I think there's a – there are big opportunities for more companies to create hyper targeted and narrow communities themselves. Right?

[00:27:42] Kevin Indig: Maybe it's around a certain use case. Maybe it's around a certain problem. Maybe it's around a certain type of person. But I really see kind of a return to not just forums, but communities. And to be honest, like, in a world where a lot of knowledge is commoditized due to AI, communities offer a space to still have secrets, at least some type of secret until it gets out.

[00:28:06] Kevin Indig: So basically, anything that is public now is immediately commoditized and available to everyone. And anything that is gated, quote unquote, and part of a secret circle, or maybe not so secret, but part of a exclusive circle, that has a lot more value. And so I think this is more than just a way to engage potential customers and to retain existing customers. I think there's also something to be said about the content you create just for your community, and the content that they create together with you.

[00:28:36] Kevin Indig: And that kind of belonging, that kind of tribalism. I see that as incredibly powerful. You can see that they've been at the space of events. A lot of companies are going back to invest more in in person events. And a big part of that is because that's where you can build real connections, real relationships, and it translates online as well.

[00:28:56] Andrew Michael: Yeah. What are these secret groups? Like, give me a few as well. Maybe we can take this offline.

[00:29:02] Kevin Indig: I can't tell you they're secret. No. I'm just kidding. Just kidding. It's less about the secrecy, but it's much more about, kind of, what is the community knowledge that we try to preserve internally and don't give away. And it doesn't – you know, like, realistically, like things will, you know, leak and maybe there aren't those massive secrets anymore. But I think the takeaway here is that when you consider or when you create a community or you already have a community, one of the stickiest ways or one of the best ways to make the community sticky is to create a body of community knowledge. Here are kind of, you know, maybe some hard earned lessons or here is what members are sharing to the community.

[00:29:45] Kevin Indig: Maybe in combination with some sort of like a code or a principle to keep that knowledge within the community, perfectly knowing that you cannot force people or assign them and have them sign an NDA. But, like, just think about the idea of being part of a circle that shares a lot and doesn't give out a lot and you kind of develop this sort of like proprietary body of knowledge. That is incredibly valuable. I see the puck going a little bit into that direction, at least in combination with a lot of the other traffic channels that we use to still get attention.

[00:30:16] Andrew Michael: Very nice. I think it's super interesting, like, where everything's heading now. So, obviously, we talked a little bit about, like, the impact on conversion rates as well when it comes to, like, searches coming through ChatGPT and stuff. I think one of the other sort of hypothesis that I would have then as well is, like, it's still too early probably to see this data, but how this plays out in churn and retention from different channels.

[00:30:37] Andrew Michael: So obviously, like, I think community, a great driver for retention. But I also wonder if, like, you tend to get a lot of churn because as well sometimes, like, marketing will oversell the product and then the product will under deliver and then, like, there's this mix match between, like, the messaging thing. Whereas, like, it feels almost now that it's, like, through ChatGPT searches, like, that gets eliminated a bit and you get to, like, the real the truth. So a hypothesis, I wonder if we start to see, like, much stronger retention rates as well coming from these generative searches over time. Maybe you've seen this already.

[00:31:06] Kevin Indig: Yep. Yeah. I don't have exact data to share. This is all very kind of spotty and anecdotal. But, yeah, absolutely. I see the traffic that still comes being, like, more willing to sign up and pay. And yet at the same time, people get so much information that they're also more likely to retain and less likely to churn. And a success criteria, the companies that are doing this the best or that benefit the most from this, they are putting a lot of investment into their documentation.

[00:31:36] Kevin Indig: So this might be a bit surprising, but help documentation and kind of tutorials and guides on all that kind of stuff, it helps – I mean, specifically for a product, but it helps a lot with AI giving better answer to the user. So if I'm evaluating a CRM, to go back to the previous example, and you are very diligent and transparent in your help and support documentation about how things work, right? Like not just API, but like anything, any feature, you give a lot of food for the AI to answer any user question correctly.

[00:32:10] Kevin Indig: And that is a gap that nobody else will fill for you, right? Because nobody else is going to create content about your health documentation. Maybe there are some mentions in reviews about your solution, but this is usually content that is hard to make explicit and get out of the software. So I see a big opportunity there. And companies that are benefiting from AI traffic, they're usually really good at explaining all the features, how they play together, what's possible and not possible in your solution. And they also do a really good job in getting feedback from users to understand what they like, what they expect, and where they might be unhappy to then reflect that back to some degree in the content that they create around their product.

[00:32:53] Andrew Michael: Yeah. 100%. I was literally having this conversation, like, two days ago, and I think it's, like, documentation now as well, like, it's the best way as well for LLMs to be able to on support, I think, from, like, a support perspective as well is, like, if you have really well defined documentation, like, it reduces the reliance as well on, like, growing support's headcount. But then, as you say as well, it has this magnifying impact on SEO.

[00:33:16] Andrew Michael: And, actually, like, I think for me, probably the company that's, like, setting the bar for this is PostHog. If you're familiar, like, we had previously James Hawkins on the show, but their, like, documentation is, like, next level, how good it is. And the whole company is basically transparent, so they have their, like, internal company docs and what and everything as well. But I think, like, that also feels like one of the new sort of, like, roles or spaces that's going to shift and change now in this next wave because it becomes critical to have this, like, really well curated docs to have access to be able to answer questions. Yeah. I mean, I'm excited to see, like, how the space will evolve as well.

[00:33:57] Andrew Michael: So we have a couple of questions that I ask every guest I wanna make sure I ask you before we wrap up today. What's one thing that you know today about churn and retention that you wish you knew when you got started with your career?

[00:34:06] Kevin Indig: The power of profiling people. What I mean is – maybe qualifying is a better word than profiling. What I mean is, for example, for an email newsletter, I think, quote, unquote, lazy execution could be, hey, sign up for our email newsletter and you ask people for maybe their name and email. I think really good execution could look like a little survey, almost like an onboarding survey of, you know, what's your role? Who do you work for? What problems do you try to solve? What piqued your interest? Etcetera, etcetera.

[00:34:35] Kevin Indig: So that you can funnel them into nurturing sequences that make sense for them. Or maybe also just tell them, you know what? You might not be, you know, like, we might not be the right solution for you. You know, I think that's totally fair to say. But the better you qualify people before you kind of send them into retention channels, like email or even evaluate your churn rate is how many of these users or subscribers actually fit our ICP and are actually a good fit for us. I think it can be very sloppy with churn in general or you can focus the numbers on the people that you actually care about.

[00:35:11] Andrew Michael: Yep. Absolutely. I think it's definitely one of those things like, especially for early stage startups. A little bit harder to do and people feel like they're missing out by focusing on an ICP and not a customer. But it's almost inevitably, like, where everybody ends up and they come to realization that's, like, we're far better off servicing, like, an audience we know that the product is ideal for in our space than trying to please everybody and then end up doing nothing. And yeah, I think, again, having those talented experiences is amazing.

[00:35:37] Andrew Michael: Last question then, and maybe we'll take this in the context of, like, SEO because I think that's where your vast majority of experiences. What's one question that you wish more people would ask you, but they don't when it comes to SEO?

[00:35:53] Kevin Indig: How can we better understand what users are really searching? I see this default in SEO to prioritize anything really by search volume. And search volume is this metric that we're getting from Google specifically from their keyword planner, which is mainly used for advertising. And on the one hand, it's the best quantification of demand and a market that we have in marketing period. On the other hand, it is a very fallible number that is very generalized, very broad. It's not super accurate and it's missing for lots of searches or topics that people care about.

[00:36:35] Kevin Indig: And so, what I wish more companies would do is to find qualitative ways to decide where to invest in SEO. Meaning, talk to your customers. Maybe, you know, like develop some sort of a program where you have a regular call with a customer every week. So, like, this is really what I'm working on with my clients is that stand up a program to talk to customers on a regular basis and truly understand what they care about, the problems they search for, how they use search, and let that inform what content we create.

[00:37:07] Kevin Indig: It's much higher signal. It's much more accurate. I've done this with a couple of companies. We see way higher conversion rates, much better retention, much more success in SEO. But obviously, it's a bit tedious. Right? It's much easier to turn on a bunch of tools, use a bunch of spreadsheets, and, like, have this refined formula to decide what to go after. But I will say that AI allows us to evaluate these conversations much, much better and draw better insights out of them. So I would like to push more companies to talk to the customers and let that drive their investment in SEO.

[00:37:40] Andrew Michael: It's interesting because it's always, like, one of the most, like, commonly cited answers for pretty much anything when it comes to, like, startups and company building and stuff is, like, speak to your customers no matter what it is. If it's, like, SEO, if it's driving growth, it's fixing retention, if it's, like – it all comes back to this. And I think, like, as you said, like, really understanding what people are searching for allows you just do so much better job with the content that you produce and how you serve it. And it almost feels like a no brainer, but like you said, people almost always default to, like, the easy option, the analytics, the data, and not getting to the truths of, like, what is the reasons behind those searches. And I think that's where, like, impactful stuff comes.

[00:38:20] Andrew Michael: Kevin, it's been an absolute pleasure chatting to you today. Is there any final thoughts you wanna leave the listeners with? Like, anything they should be aware of to keep you up to speed with your work?

[00:38:28] Kevin Indig: Yeah. Thanks so much, Andrew. This was a lot of fun. Great questions. You know, one thing to point out maybe is this usability study that I recently published. You can find that on growth-memo.com. I publish a lot of original research there on a regular basis, so go check it out. You know, if you don't like it, leave me some feedback. I always wanna understand how to do a better job. Or connect with me on LinkedIn, Kevin Indig. You should find me relatively quickly. But that's really it. Thanks so much for having me, Andrew. This was fun.

[00:38:55] Andrew Michael: Amazing. For the listeners, we'll make sure to leave everything we discussed today in the show notes so you can pick that up there as well. And thanks again for joining, Kevin. I wish you best of luck going forward.

[00:39:03] Kevin Indig: Thanks, Same to you.

[00:39:04] Andrew Michael: Cheers.

[00:39:05] Andrew Michael: And that's a wrap for the show today with me, Andrew Michael. I really hope you enjoyed it and you're able to pull out something valuable for your business. To keep up to date with Churn.FM and be notified about new episodes, blog posts, and more, subscribe to our mailing list by visiting churn.fm.

[00:39:25] Andrew Michael: Also, don't forget to subscribe to our show on iTunes, Google Play, or wherever you listen to your podcasts. If you have any feedback, good or bad, I would love to hear from you, and you can provide your blunt direct feedback by sending it to andrew@churn.fm. Lastly, but most importantly, if you enjoyed this episode, please share it and leave a review as it really helps get the word out and grow the community. Thanks again for listening. See you again next week.